Storing OpenAI embeddings in Postgres with pgvector

Re:Storing OpenAI embeddings in Postgres with pgvector by visarga on Monday February 06, 2023 @05:48PM (#34685628)

Played with GPT-3 embeds for the first time this week and they seem to be unusually good. Everything >0.85 cosine similarity is a match, everything under 0.76 is not a match, and in the 0.76 .. 0.85 is the boundary zone where some positives and negatives get mixed up. Training models on top works great, even averaging 2-3 phrases works great for making a more precise query.

[ Parent | on HN ]

Re:Storing OpenAI embeddings in Postgres with pgvector by kanyethegreat on Monday February 06, 2023 @07:09PM (#34686524)

> Training models on top works great, even averaging 2-3 phrases works great for making a more precise query.

What does this mean?

[ Parent | on HN ]

Re:Storing OpenAI embeddings in Postgres with pgvector by visarga on Monday February 06, 2023 @07:12PM (#34686553)

If you average a query embedding with that of similar terms, you get a more focused ranking. Averaging amplifies the common component of the vectors and diminishes the spurious components.

[ Parent | on HN ]

Re:Storing OpenAI embeddings in Postgres with pgvector by chipgap98 on Monday February 06, 2023 @07:19PM (#34686614)

What do you mean by "average a query embedding"?

[ Parent | on HN ]

Re:Storing OpenAI embeddings in Postgres with pgvector by CardenB on Monday February 06, 2023 @08:03PM (#34687062)

When you input a query, you get a numerical representation output. If you have 3-5 similar queries (each phrased a bit differently), they all have slightly different outputs. If you average them all together, it combines them in a way that improves results.

[ Parent | on HN ]

Re:Storing OpenAI embeddings in Postgres with pgvector by teaearlgraycold on Monday February 06, 2023 @08:27PM (#34687258)

Presumably this operation:

  def average_embeddings(embeddings):
    return [
      sum([embedding[x] for embedding in embeddings]) / len(embeddings)
      for x in range(len(embeddings))
    ]

Just finding the mid-point in N-dimensional space between all points. Although I believe OpenAI says these are directions rather than locations. Averaging makes sure you don't accidentally pick a target vector that happens to be on the fringes of your desired embedding space.

So if you're querying a document for predatory birds and you are using the string "Eagles" as your target you will be accidentally digging into embedding space for sports teams (go birds). But if your target is `average_embeddings([eagles_embedding, hawks_embedding, falcons_embedding])` you'll end up with less noise.

[ Parent | on HN ]

Re:Storing OpenAI embeddings in Postgres with pgvector by crosen99 on Monday February 06, 2023 @07:30PM (#34686723)

That’s good info. Curious if you have any guidance about how big a document chunk should be to get a good match with a natural language query. For context, I’m building a question answering app using ada-02 that pulls in around 100 pages of text.

[ Parent | on HN ]